The debate on the competitive risks of common ownership has focused on whether passive index investments soften competition among portfolio companies. If this were the case, it would raise the question how it would impact the analysis of horizontal mergers between portfolio companies. The European Commission has, in Dow/DuPont and in Bayer/Monsanto, allowed the possibility that the mere existence of common ownership can contribute to a competitive impediment emerging from a horizontal merger between portfolio companies.
This paper, available here, argues that it should not be presumed that common ownership in itself increases anticompetitive effects of a merger between portfolio companies. Instead, this would depend on the facts of the case – from both a price and innovation perspective. Even if it were to be assumed that common ownership has an inherent propensity to impede competition, it cannot be concluded from this that a merger between commonly owned portfolio companies has a larger detrimental effect on prices or innovation activity than would have been the case absent common ownership. Without a thorough analysis of the means of minority influence in a given case, of the shareholder structure on the affected market and on adjacent markets, and of potential efficiencies arising from the merger, it is not possible to conclude from common ownership on the likeliness of competitive harm to result from the considered merger.
Sections II A and B define common ownership and assess the practical relevance of common ownership based theories of harm for horizontal mergers.
Common ownership occurs when the same investor(s) hold(s) minority stakes/make portfolio investments in multiple competitors. The economic hypothesis behind the alleged harmfulness of common ownership is that of an indirect internalisation of external effects of price increases on their common shareholders. A horizontal merger between portfolio companies therefore basically raises two questions with respect to possible impacts of common ownership on the counterfactual analysis: (a) to what extent did the firms internalised pricing externalities due to common ownership before the merger, and (a) whether and to what extent does a horizontal merger increase the degree of this internalisation of pricing externalities?
While this is predominantly still an academic debate, the European Commission has referred to the potential harmfulness of common ownership in the context of horizontal mergers between portfolio companies in a couple of recent merger control decisions. Although the existence of common ownership does not seem to have been dispositive in any of these cases, they nonetheless present a theory of harm based on the hypothesis that a horizontal merger on a market where common ownership is an issue can lead to an even greater harm to effective competition than what would occur ceteris paribus absent common ownership.
The authors contest this approach. Any horizontal merger, by definition, leads – to some extent – to an internalisation of external price effects. However, depending on how investors have spread their holdings on the market, a merger that directly affects the structure of common ownership, and thereby firms’ incentives to set prices, need not necessarily create upward pricing pressures. The mere fact that common shareholdings exist is in itself insufficient to argue that competitive harm is more likely or serious in the event of a merger. Instead, it must be shown that the channels through which common ownership can have an effect on firms’ strategies are present, and integrate this demonstration into the wider merger analysis.
Section II C shows that common ownership may mitigate post-merger price effects.
Assuming that common ownership has an impact on firms’ strategic behaviour, there are a number of ways through which common ownership can modify the effects of a merger. Some recent contributions have focused on the use of the modified HHI (M-HHI) as a way to integrate common ownership into a single concentration measure. The problem with such approaches is that concentration measures are often insufficient to draw conclusions on specific merger effects. Moreover, unlike HHI, a M-HHI is unlikely to prove useful as a screening tool, since the overall effect of common ownership on individual firms’ price-setting incentives, pre- and post-merger, is not susceptible of being caught by such a tool. However, the derivation of the upward pricing pressure (or of other ‘pricing pressure indices (PPIs)’) can more directly measure the effects of a merger. Once the way in which common ownership affects a firms’ pricing incentives has been pinned down, integrating common ownership into UPP measures is straightforward. However, this has limitations. While UPP measures hold the pricing by competitors of the merged firm constant, with common ownership such ‘second round’ price effects may however be more pervasive since the incentives of parties to the merger will now potentially incorporate also the margins and profits of outsiders.
Given this, the authors confine themselves to illustrating how common ownership may affect the incremental internalisation of former rivals’ profits when comparing a firm’s pre- and post-merger pricing incentives. To do this, they develop a model where two investors hold stakes in competitors, and where the management places on each owner’s profits some weight on each investor’s profits. With premerger common ownership, the management of firm 1 already internalises the effects of its strategy choices on the rival 2 before the merger, implying, at least, that the incremental internalisation effect from the merger is smaller in the presence of common ownership. Ceteris paribus, this means that, under certain circumstances, the upward pricing pressure from the merger could be mitigated by the presence of common ownership.
In effect, in specific circumstances and in the presence of pre-merger common ownership, a merger can even have the somewhat perverse effect of reducing the internalisation factor so that the merger would instead have a tendency to exert a downward and not an upward pricing pressure on the respective firm. In particular, this would be the case if, as a result of common ownership, the management of firm 1 would actually place more weight on the rival’s profits than on that of firm 1 before the merger. This might be the case if management of firm 1 cares more about the profits of some owner A who, at the same time, has a higher financial interest in the rival firm than in firm 1. The authors acknowledge that such an outcome may seem unrealistic, and it could amount to a breach of the fiduciary duty of firm 1’s management; but it shows how the effects of common ownership are not all in the same direction in all scenarios. One can thus conclude that pre-merger common ownership may mitigate the price effect resulting from a merger even in the absence of efficiencies.
Section II D looks at how common ownership may impact innovation after a merger.
In Dow/DuPont, the Commission held that the alleged harmfulness of common ownership in a horizontal merger context is not confined to price effects, but that the underlying theoretical framework is supposed to be transferable to the analysis of innovation competition. The authors argue that the same framework of analysis they developed as regards pricing shows that the effects of common ownership on the innovation effects of a merger can go either way. In other words, a merger’s dampening effect on innovation may again be lower under pre-existing common ownership, given that the negative effect of such business stealing may already have been partially internalized before. Once again, the analysis of how common ownership bears on the effect of a merger on innovation needs a proper assessment, provided that one presumes that common ownership is of relevance at all in a specific case.
Section II E looks into the standard of proof for theories of harm based on common ownership.
The Commission carries the burden of proof for the anticompetitive effects of a merger. When it comes to relying on common ownership as an element of a theory of harm, it is therefore upon the authority to adduce evidence that support its allegations. That is where the intricacies of the underlying theoretical framework of common ownership bulk up. Given its ambiguous effects, lack of evidence cannot be bypassed by presumptions flowing from common ownership to the detriment of the parties. To some degree, parties can be retained to support the authority in its investigations by providing necessary information which is in their possession. Yet this does not exonerate the authority from its duty to autonomously gather information from other sources, or from its burden of persuasion.
In the light of these stipulations, the antitrust authority should be retained to demonstrate that, based on the available data, the assumption of minority influence by common owners is corroborated. Prima facie evidence would require that, upon initial examination of the case, sufficient corroborating evidence appeared to exist in support of the allegation. The mere existence of common ownership, however, is inapt to supplant such assumption. In effect, there are many scenarios where common ownership may not create incentives to engage in anticompetitive conduct – when fiduciary duties are observed, institutional investors may have different balances of investments across portfolio firms and favour the profitability of different firms in an industry, etc. The authors therefore conclude that any assumptions with respect to the anticompetitive effects of common ownership in horizontal merger contexts hinge on basically two limbs: (i) the antitrust authority must demonstrate that an internalisation of the common shareholders’ interests actually takes place; and (ii) that the merger alters the determinants of internalisation in a way which induces unilateral harmful conduct resulting from the merger.
This paper is a thoughtful analytical piece that provides a different perspective on common ownership. I really enjoyed how the paper effectively mixed economic and legal analysis, which is particularly important for the assessment of factual assumptions and presumptions in competition law.
I am broadly sympathetic to the analysis, and would be surprised if a presumption of harm could be derived from the pre-existence of common ownership among merging firms. At the same time, I was not fully persuaded by the paper’s arguments.
First, the paper goes to great lengths to demonstrate that common ownership pre-merger can reduce the anticompetitive effects of a merger. However, when all is said and done, the scenario that the authors were able to come up with where common ownership can have pro-competitive effects is marginal. The authors themselves explicitly recognise the unlikelihood of such a scenario, and then devote a full section to explaining how ‘It would, however, be wrong to conclude that pre-merger common ownership will always mitigate the price effects of a merger. In fact, (…) common ownership may well reinforce the upward pricing pressure that generally results from the internalization effects of a merger.’ This is a tribute to their intellectual honesty, but it does undercut their argument somewhat.
I was also slightly confused by the last section. While the discussion on evidentiary rules is impeccable, it seems to focus on how to prove whether common ownership is able to influence corporate behaviour. However, the earlier economic section of the papers showed that common ownership could have ambiguous effects, but always assuming that common ownership is able to influence corporate behaviour.
To my mind, the most important question, even from a legal/evidentiary standpoint, concerns whether we can establish the effects that common ownership will have on competition, not whether one can prove a causal pathway between common ownership and corporate behaviour. As noted in a paper I discussed last week, it is generally accepted that certain types of common ownership can have effects on corporate behaviour. Questions concerning how this occurs are, however, mainly relevant to determine whether intervention is justified, and what form it should take.
In any event, without an economic consensus that common ownership can influence corporate behaviour, competition enforcement in this area is extremely unlikely. If such an economic consensus emerges, the relevant question in an enforcement context will likely focus on whether it can be proved in each case that common ownership’s influence over corporate behaviour works to augment or reduce price or innovation internalisation – or has no competitive effects at all. In short, it strikes me that the discussion in the last section, on whether common ownership can influence corporate behaviour, would be moot in most cases. If enforcement were to occur, it would already be generally assumed that common ownership could influence corporate behaviour, and the evidentiary analysis would likely focus on the nature of the competitive effects of common ownership on a merger – which was the focus of the economic analysis in this paper.
Further, reading the paper I got the distinct sense that common ownership would, except in very narrow circumstances, tend to produce anticompetitive outcomes. Given this, it is at least arguable that a presumption of likely harm to competition could be justified for some mergers involving common ownership. Of course, such a presumption would require that a number of things be established beforehand, e.g.: that common ownership be shown to be able to influence corporate behaviour in general; that a specific merger falls within one of the situations where there is evidence of common ownership typically affecting corporate behaviour; and that the merger does not fall within the small category of transactions where common ownership can reduce internalisation incentives.